A question that’s been asked many times. The answer is complicated, but the simplest way it to put it, is…. it depends on what you mean by good.
We could define “good” as wanting the music we are listening to sound the same as it did when it was recorded. If we ignore all of the intervening steps from a recording studio to the streaming digital that you are most likely to be listening to on your motorcycle, the part that Audyse has control of at least is the headset sound reaching your ears.
Traditionally, the performance of a loudspeaker and headphone (smaller speaker in a small space) is measured by how “flat” the response of the speaker is across the audio range, 20 Hz to 20 kHz. In Figure 1 below, we have a hypothetical headphone response, which is the black solid line.
In order to understand this plot, we need to define what we mean by “Speaker Response” (the vertical axis) in the plot. Zero dB is the ideal response for a speaker, because this means that the speaker neither emphasizes (amplifies) nor deemphasizes (reduces) the sound more than the original recording. And for “perfect” sound reproduction, we want the speaker to have zero dB across its entire range of frequencies (the Green Line in the above plot). Since this is practically impossible to achieve in reality, we can say that it is good enough if we stay within plus and minus 3 dB of zero (the red dashed lines).
So, to interpret this plot, we can see that between 20 Hz and about 35 Hz (first pink region), the speaker is producing sounds that are quieter than would be in a music recording. From 35 Hz to around 150 Hz (first blue region), the speaker is producing sounds a little louder than the music recording, but it is within the allowable 3 dB amount. And so on for the other 2 regions (one blue, one pink).
So the plot is cool, but what does it mean to what we hear? These response variations at different frequencies can make your sound “thin” (lacking in bass) or “muddy” (lacking in highs), “boomy” (too much bass), etc.
To correct for these variations in the frequency response of a speaker, we can apply EQ (short for equalization) to the drive signal being sent to the speaker. EQ selectively applies gain (amplification) or reduction to the signal only in the appropriate frequency ranges in order to achieve a more balanced, or “correct” sound. And, if we were to repeat the frequency response measurement, we get the plot in Figure 2.
The black line has been shifted up and down bringing it more in line with the “perfect” music reproduction”. In practice, EQ can only correct for some variations, but this will sound “better” than the non-equalized speaker.
Audyse uses Digital Signal Processing (DSP) technology to produce a flat response in our headset, giving you an audiophile listening experience. Since speaker drivers have unique properties, we measure and calibrate the headset EQ as we are manufacturing them.
There are many other technical factors that can influence “perfect sound reproduction”, including distortion (non-linearity), signal to noise of the amplifier, digital compression, etc., but we will leave these issues for another blog!
Ultimately, all sound reproduction is a compromise, and what determines “good” sound – once basic EQ has been applied – is a matter of personal taste. The EQ that you prefer for listening to Rock music might be completely different from what someone else prefers to listen to for Classical music or Hip-Hop. And these are subjective choices.
